Test Methods to Characterize Participate Matter Emissions
and Deposition Rates in a Research House
Jenia McBrian and Roy Fort man n
ARCADIS Geraghty & Miller, Inc., P.O. Box 13109, Research Triangle Park, NC 27709
and
Zhishi Guo and Ronald B. Mosley
U.S. Environmental Protection Agency, Office of Research and Development, National Risk
Management Research Laboratory, Research Triangle Park, NC 27711
ABSTRACT
There are a number of potential sources of indoor particulate matter (PM), including combustion
appliances, cooking, vacuuming, and various other human activities. Because of potential health
hazards associated with exposure to fine PM, it is important to characterize indoor PM sources
and the fate of the particles emitted from those sources. Test methods have been developed to
measure PM mass emission rates and particle size distributions and to estimate PM deposition
rates under close-to-realistic conditions in a research house.
One room in the research house has been specially configured for PM source testing. A high-
efficiency particle air (HEPA)-filtered air supply system, used for positive pressurization of the
room, minimizes PM infiltration from the outdoors and the rest of the house while a portable
HEPA filter in the room reduces background PM prior to the start of source tests. A ceiling fen
and portable air conditioner provide mixing and temperature control Testing involves
simultaneous real-time particle monitoring and integrated sampling. Particulate matter with
aerodynamic diameters less than 2.5 and 10 um (PMa.s and PMio, respectively) are collected on
Teflon filters using size-selective cyclones operated at 16.7 L/min. Particle size distributions are
measured with an Electrical Low Pressure Impactor (ELPI) with 12 size fractions, from 0.030 to
10 um aerodynamic diameter. Other parameters measured during tests include temperature,
relative humidity, air exchange rate, and outdoor meteorological parameters. During tests with
combustion sources such as kerosene heaters, carbon monoxide, nitrogen oxides, and sulfur
dioxide are also measured.
These test methods have been used to characterize emissions from an unvented gas space heater,
kerosene heaters, candles, and incense. This paper describes the sampling methods, measurement
parameters, and technical approach for PM source tests. The utility of this technical approach to
measure emission rates and estimate deposition velocity is demonstrated by presentation of results
obtained from tests with a kerosene heater and incense.
INTRODUCTION
Research by the U.S. Environmental Protection Agency (EPA) suggests that exposure to
particulate matter (PM) can constitute a potential health hazard. Because of concern about
exposure to PM, the EPA has developed National Ambient Air Quality Standards for particles
with aerodynamic diameters less than 2.5 u,m (PMz.s) and 10 (im (PMio). Although outdoor PM is
important, recent studies suggest that indoor sources may account for a substantial fraction of the
-------
and PM2.s measured indoors. While there has been extensive research on outdoor sources of
PM, indoor sources remain largely uncharacterized.
Previous methods used to test indoor combustion sources include a large environmental chamber,1
a sampling hood and duct,2 and a test house without controls to minimize background particles.3
Small chambers have often been used to test small sources such as candles.4'5 The method
described in this paper allows the indoor combustion sources to be tested under close-to-realistic
conditions while particle infiltration from outdoors and adjacent rooms is minimized. This method
also allows for the determination of PM emission and deposition rates in a single experiment.
Knowledge of PM deposition rates allows for compensation of wall loss when computing the
emission rate.
To characterize PM emissions from indoor sources and to estimate PM deposition rates, one room
of a research house has been configured with a system to minimize particle background
concentrations during tests. Positive pressurization of the room from a high-efficiency particle air
(HEPA)-filtered air supply system minimizes PM infiltration from the outdoors and the rest of the
house. A portable HEPA filter further reduces background PM in the room prior to the start of
source tests. A ceiling fen and portable air conditioner provide mixing and temperature control.
Particle concentration and size distribution are measured with a real-time particle monitor.
Integrated samples of particles are collected on filters using size selective samplers for PM2 5 and
PMio size fractions. Other parameters, including nitric oxide (NO), nitrogen dioxide (NO2),
carbon monoxide (CO), and sulfur dioxide (SOa) are measured continuously with combustion
appliances (space heaters). The data will be used to estimate emission factors and particle
deposition velocity. This paper describes the setup of the test room, test methods, and gives
examples of test results.
EXPERIMENTAL
Indoor Air Quality (IAQ) Research House
Tests were conducted in one bedroom of the EPA Indoor Environment Management Branch
(IEMB) IAQ research house. The house is an unoccupied one-story ranch-style single-family
residential dwelling of standard wood frame construction. The floor plan consists of a kitchen,
dining room, living room, and den at one end of the house, and three bedrooms and two
bathrooms at the other. The floor area is approximately 121 m2 (1300 ft2) and the volume is 300
m3. The house is fully equipped with instruments for measuring environmental parameters, air
exchange rates, gaseous pollutants, and PM. The attached garage serves as the laboratory for the
research facility.
Configuration of the Test Room
The experiments described in this paper were performed in the front corner bedroom (FCBR) of
the research house (see Figure 1). The room has dimensions of 3.78 (length) x 3.28 (width) x
2.44 (height) m and has a volume of 30.2 m3, not including the closet, which was sealed prior to
testing. It has vinyl flooring, painted gypsum walls, and a textured gypsum ceiling. The room was
isolated from the main house heating and air-conditioning (HAC) system for these tests.
-------
Figure 1. FCBR diagram.
ll
i n
i o
! i Window
j ;
Fans
Air"^ -"
supply :
HEPA
, CEM
'\ ' port
Pilot SF* dosing
Adjoining
bedroom
i
\
WJ
CEI»
inl<
/
/
N2
\
(
r-v
A
;t \
TableV-
iELPIl
/ \
Front window
i Source,-
Veiling;' i
"""'"-, ' '
/ / ' \
\\
i \ "N
o| \\\
\ \ ''. \ x-
\\\ FCBR'-\
^N Vu
W3
B&K
sampling
port °
6 RH /-
Pressure
transducer
^
o
o
C '
-Cyclones 3^
1
I
i AC
: unit
f
Ceiling
fan
B&K
sampling port
PLAN VIEW
] ELPI and cyclone
j pumps (located
; outside of room)
,
; Closet
; , ...
(sealed)
Temp/;' ,. Cyclones ;
^°RH i!/.-^- ;
transducer
Source
in,et
W2
W4
d.
U
as
c
CEM
inlet -
(hidden)
sil
fai
"g
1(
prf ,
' I Front window
1
1 ELPI
Wl
|
; HEl'A
i ! 1
! i
i ;
| 1
i <
n n
v '^
\
Source
1
1
Temp/ J
RH ^,/-t
Pressure ___»^Q
- transducer j!
f - Cyclones 1;
"5;
s)
J/5
W3
B&K
t- sampling
port
AC
unit j
i i
SIDE VIEW
END VIEW
PROTECTED UNDER INTERNATIONAL COPYRIGHT
ALL RIGHTS RESERVED
NATIONAL TECHNICAL INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
Reproduced from
best available copy.
-------
There are two windows in the FCBR, as shown in Figure 1. A window air-conditioning (AC) unit
is mounted in one of the windows, 0.7 m above the floor. The maximum output of this AC unit is
7,700 Btu/hr (1940.4 keal/hr). When used during testing of large combustion sources, the unit is
operated without the filter on the high (maximum cool) setting.
To promote good air mixing in the room, a low-profile five-blade ceiling fen is located near the
center of the room. The fen bkdes span 1.32 m, tip to tip, and hang 0.24 m from the ceiling. The
fen is operated at various settings, depending on the test being conducted. During tests with space
heaters, the ceiling fan was operated on low, with the fan blades moving air down. During tests
with candles, the fan speed and direction was varied to examine the effects of airflow on PM
emission rates. For incense tests, the fen was operated on low with the airflow in the downward
direction.
Prior to testing, the FCBR was configured with an air supply system to facilitate positive
pressurization and to minimize the infiltration of particles from the outdoors and other parts of the
house. Major penetration areas around windows, doors, electrical outlets, and baseboards were
sealed, and differential pressure sensors were installed to monitor the pressure between the FCBR
and outdoors. A constant dosing experiment with sulfur hexafluoride (SFs) was performed to
evaluate the pressurization system. Differential pressures are monitored during each experiment.
The pressurization system is operated at a flow rate of approximately 53 m /hr to achieve a
pressure of 2 to 4 Pa relative to the outdoors.
Particle-free air is supplied to the room through a 0.63 m ID polyvinyl chloride (PVC) pipe located
in wall 1 (Wl), 0.61 m from the ceiling. The outlet of the pipe is angled towards the ceiling to
minimize interference with the source. The air supply through the pipe is generated by two in-line
fens (FanTeck, Model FR250,2400 rpm max) which draw outdoor air through an in-line HEPA
filter (SAFEMOD, Model 6PH1000). Fan speed is controlled by a variable autotransformer. A
calibrated pilot tube is mounted in the pipe, from which readings are taken manually at 15-minute
intervals using an Airdata Multimeter (Model ADM-850).
Pressure differentials are measured between the FCBR and the outdoors. The outdoor pressure is
measured through two joined tubes extending at least 1 m from the two exterior walls (W2 and
W4) of the room, A calibrated pressure transducer (Validyne Engineering Corp., Model DP851V-
PO5) with a range of 0 to 0.050 in. HzO and an output of 0 to 5 VDC is used for continuous
pressure measurements. Output voltage is recorded with a data acquisition system (DAS), located
in the garage.
A portable HEPA filtration unit (Bionaire, Model CH-3580 or Honeywell Enviracaire, Model
13520) is located in one corner of the room, 0.43 m away from both Wl and W2. This unit is
operated in the room prior to each test to remove particles resuspended during setup and test
preparation.
A temperature/ relative humidity (RH) sensor (HyCal Sensing Products, Model HHT-2WC-D9-
TTA), located 0.66 m from the ceiling, is used to record data at 1-minute intervals with the DAS.
-------
For air exchange measurements, SFe is injected into the room through a port on the bottom of the
pressurization pipe, in the adjoining bedroom. Air exchange measurements are made with a Bruel
& Kjaar (B&K) Type 1302 Multi-Gas Monitor using a SFa tracer gas decay method based on
ASTM E741-95.6 This decay data, in tandem with particle data recorded by the ELPI, is used to
estimate particle deposition rates. The sample is drawn through a polytetrafluoroethylene (PTFE)
filter and delivered to the instrument through polyethylene tubing using a B&K 1303 multipoint
sampler. This sampling inlet is suspended 0.76 m from the ceiling, at the location indicated in
Figure 1.
Source Locations
During testing, sources are located approximately 1.32 m from Wl and 0.61 m from W2, as
shown in Figure 1. Incense and candle sources are elevated to a height of approximately 1.20 m,
and space heaters are operated on the floor, consistent with normal use.
Measurement Parameters and Methods
Tests with combustion sources involve measuring numerous parameters. The parameters and
methods are summarized in Table 1. Integrated samples of PMj.s and PMio are collected on 47
mm, 2 um Teflon filters (Pall Corp., Teflo, Part No. R2PJ047) using University Research Glass
(URG) cyclones (URG-2000-30EH, 2.5 um and URG-2000-ENB, 10 urn), operated at a nominal
flow rate of 16.7 L/min. Cyclone inlets are located 1.57 m above the floor. PM2.5 and PMio
samples are collected in duplicate for most tests. PM mass collected on filters is subject to
gravimetric analysis after sample collection. Gravimetric analyses are performed in a controlled
environment weigh room. Filters are conditioned for 24 hours at 35% RH prior to weighing.
During all tests, a DEKATI ELPI (distributed by TSI Particle Instrument Division, Shoreview,
MN) real-time particle monitor continuously records 60 second averages of aerosol particle
concentrations in 12 size fractions. The instrument range is from 0.03 to 10 um aerodynamic
diameter. Data collected by the ELPI is used to estimate PM emission rates from each source type
and to estimate deposition velocity of particles in different size fractions. The instrument is
located adjacent to the integrated samplers (Figure 1) with the inlet 1.65 m above the floor.
Gravimetric mass data were compared to those collected by the ELPI. These data are presented in
a paper presented in these proceedings.7
During the space heater tests, gaseous pollutants are continuously monitored with an SOj analyzer
(Thermo Environmental Instrument (TEI), Inc., Model 43A), a nitrogen oxides (NOX) analyzer
(TEI, Model 46), and a CO monitor (TEI, Model 48). All three instruments are continuous
emissions monitors (CEMs), Samples are collected from the FCBR and the air supply system
through Teflon sample lines and a pump with a Teflon-coated diaphragm. The FCBR sampling Ike
is located 1.57m above the floor. Samples from the pressurization tube are collected from a port
located on top of the pipe, 0.23 m from the wall where it enters the room. This sampling location
ensures that the concentrations measured represent the air delivered to the room following passage
through the fans, HEPA filter, and tube, not the outdoor concentration, which may be different.
An automated system switches sample collection between the two sampling locations every 10
minutes. Pollutant concentrations are recorded by the DAS as voltage output and are later
-------
converted to concentration units using the calibration equation associated with the relevant
instrument.
Table 1. Measurement parameters and methods.
Measurement Parameter
Meteorological (Temperature,
relative humidity (RH),
barometric pressure, wind speed
and direction)
Indoor Temperature
RH
Air Exchange
PM Concentrations and Particle
Size Distributions
PMio & PM2,5 Mass
Estimated Deposition Velocity
Nitrogen Oxides (NOX), NO, NO2
SO2
CO
Source Mass
Sampling/Analysis Methods
Continuous
Resistive temperature device
Thin-film capacitance sensor
SFg tracer gas decay
Continuous, 60 second
averages
Integrated-gravimetric
Continuous, 60 second
averages
Chemiluminescence monitor
Ultraviolet detector
Gas filter correlation monitor
Gravimetric
Instrumentation
Vaisala
HyCal Sensing Products
HyCal Sensing Products
B&K 1302 Multi-Gas
Analyzer
ELPI
URG Cyclone/filter
ELPI and B&K 1302
Multi-Gas Analyzer
TEI Model 46
TEI Model 48
TEI Model 43A
Micro Balance
Test Protocols
Test protocols are similar for all sources. The primary variable is the duration of the test, which is
a function of the PM emission rate and the time required to collect sufficient mass by the
integrated sampling method for gravimetric analysis.
The test protocols involve the following primary activities:
Start the outdoor air supply system to pressurize the room to approximately 2 to 4 Pa relative
to the outdoors.
* Close the test room and operate the air cleaner to reduce background PM concentrations.
» Start operation of the ELPI to measure room air PM background concentrations.
» Zero and span the pollutant monitors.
» Prepare the combustion source and obtain tare weights for the fuel source.
» Set up the integrated samplers in the room,
» After all setup of source and samplers is completed in the room, close the door, and operate
the air cleaner for an additional 30 minutes to reduce PM background levels.
After 30 minutes, turn off the air cleaner.
« Enter the room and start the source (do not start integrated samplers).
One hour after starting the source, begin integrated sampling.
Operate the source for the required duration, based on predicted PM emissions.
-------
« At the end of the test, turn off the integrated samplers and the source, and continue monitoring
with the ELPI and pollutant monitors.
« Inject SFe and measure air exchange rates overnight.
* At the end of the test, weigh the fuel source and calculate the mass of fuel used.
RESULTS AND DISCUSSION
Stability of Ventilation Rate
To determine the stability of the ventilation rate in the room, SF& was injected continuously at a
rate of 3.1 mL/min overnight while the room was sealed and the pressurization system on. Room
pressure was maintained at an average of 2 Pa during the experiment, and the air exchange rate of
the room was 1.1 h"1. These data, depicted in Figure 2, show that the ventilation rate was stable
during the period.
Figure 2. FCBR ventilation rate study.
0 1 2345 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Average SF, Concentration =4.48 pom
0.0
1 2345 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Elapsed Time (h)
PM Background
To establish background particle concentrations in the room, the ELPI is operated for about 30
minutes prior to each source test, while the room is sealed and under pressure. Figure 3 details
background particle concentrations before (1), and during (2) room cleaning, and after the start of
a source (3). After room cleaning, the background concentration maintained a level of less than 2
ug/m3.
-------
Figure 3. FCBR background.
Elapsed Time (b)
Effect of Window AC on PM Counting
A window air conditioner was required during operation of kerosene and unvented gas space
heaters. After installation, the filter was removed and the unit was operated in the recirculating
mode to help minimize particle loss. Additionally, the area around the AC unit was sealed with
caulk. A stick of long burning incense was burned in the room approximately 2 m from the AC
unit to determine the magnitude of particle loss caused by the operation of the AC during source
tests. The pressurization system was turned on and the FCBR closed. A HEPA filter was run for
30 minutes in the room to reduce background particle concentrations prior to the test.
The ELPI recorded particle concentrations in 1-minute averages throughout the test. After 30
minutes, the HEPA filter was turned off and a technician entered the room to light the incense.
After lighting, the FCBR door was again closed, and the room remained under about 2 Pa pressure
throughout the test. After approximately 1.5 hours, when the particle concentrations had reached
an apparent steady state, the AC unit was activated by a remote switch and operated for 30
minutes on the high setting. After 30 minutes, the AC unit was turned off, and the incense was
allowed to burn out completely. While the effect of the AC unit on particle counting appears
negligible, there may be an approximate 15% loss of PM mass during its operation. This is
consistent with the expectation that large particles will be lost on the surfaces of the AC unit.
Total mass concentrations during the test are presented in Figure 4.
-------
Figure 4. Effect of AC on PM counting.
0.5
1.0
1.5
2.0 2.5 3.0 3.5
Elapsed Time (h)
4.0
4.5
5.0
5.5
Calculation of Deposition Rates
ELPI and tracer gas data collected during the decay phase of each test are used to estimate PM
deposition rates. First order decay rate constants are determined by calculating the slope of the
line created by plotting the natural log of the decreasing concentration of PM or SFg against
elapsed time (see Figure 5). The size-dependent, first-order deposition rate constants are
determined by comparing the decay rates for PM and the SFe tracer gas, and are calculated from
Equation 1.
Equation 1. First order deposition rate constant
where:
Du = first-order deposition rate constant for PM of size i (h"!)
St = first-order decay rate constant for PM of size i (h"!)
SSF = first-order decay rate constant for the SFa tracer gas (h"1)
In exposure estimation and risk assessment, the fine PM is treated as a single air pollutant. Thus,
determining the decay for PM2.5 is of practical significance. It can be estimated in two ways: (1)
sum up all the stages with aerodynamic diameters less than 2.5 |im and compare the concentration
decay rate with that for the tracer gas, and (2) compute the mass mean diameter and use this data,
-------
presented in Figure 6, to determine the rate constant at the mass mean diameter by interpolation.
From data collected during a test with incense, the decay rate constant for PM^s was 0.28 h"1 from
method 1 and 0.31 h"1 from method 2.
Figure 5. PMb.s and SF$ decay rates from an incense test.
E
"S
o
a
a
o
O
Z.
0.2 0.4 0.6 0.8 1
Eiapsed Time (h)
1.2 1.4 1.6
Figure 6. Estimated PM deposition rate constants from an incense test.
4.5
3.5
2.5
1
0.5
0
0.01
X
0.1 1
Aerodynamic Diameter (um)
10
The data presented in Figures 5 and 6 are from the same incense test.
-------
Calculation of Emission Rates
From filter samples, the average PM emission rate during the sampling period can be calculated
from Equations 2 and 3. Equation 3 is preferred because it compensates for wall losses. Under
the test conditions described above, the results of Equation 2 are about 30% lower than those of
Equation 2.
Equation 2. Average emission rate from filter samples without compensation for wall losses.
where:
./? = PM emission rate (ug/h)
Q = ventilation flow rate (m3/h)
Wj- = PM mass on filter (ug)
Vf = sampling volume (m3)
Equation 3. Average emission rate from filter samples with compensation for wall losses.
wf
f
»r,
where:
V = room volume (m3)
Dk = first order deposition rate constant (h"1)
ELPI data collected during source tests are used to estimate the real time source PM emission
rates based on the following equations:
Equations 4 and 5. Calculation of time-varying emission rates,
»xf
V~- = R
dt
R.*V~
A/
-------
where:
V = room volume (m3)
C = PM concentration (fxg/m3)
t ~ time (h)
R = emission rate (ng/h)
Q = ventilation flow rate (m3/h)
Dk = first-order deposition rate constant (h"1)
The term - can be calculated numerically.8 The calculated emission rate profile from a
A/
kerosene heater test is shown in Figure 7.
Figure 7. Calculated PMa.5 emission rate profile for a kerosene heater.
3.E+04
2.E+04 -
C
l.E+04 -
O.E+00 4-
0
0.5
1.5 2 2.5
Elapsed Time (h)
3,5
CONCLUSIONS
The test protocols and methods outlined above facilitate both emission measurements of particles
emitted from combustion sources and the estimation of particle deposition rates following the
cessation of source usage. The tests are performed in a room with realistic wall, ceiling, and floor
-------
surface materials. Because a slight pressure (2 to 4 Pa) is maintained in the room during source
tests using a HEPA-filtered air supply, the infiltration of particles into the room from the outdoors
or other parts of the house is minimized. The air supply system ensures that most of the particles
in the room air during a test can be attributed to the source. The pressurization system has been
shown to be stable and effective.
For operation of large combustion sources, such as kerosene heaters, a window air conditioner is
required to handle the large heat load in the small room. Tests have shown that an approximate
15% loss in PM mass can be attributed to the continuous use of the AC unit. This loss is likely
caused by the impaction of large particles on the AC unit surface. A method of supplying cooled
air to the room using the pressurization system is being investigated.
As demonstrated by the data from incense tests, use of the SFe tracer gas decay method and the
real-time particle monitor facilitates calculation of deposition rate constants. The test room and
methods are useful for characterizing particle emissions from both large (e.g., kerosene heaters)
and small (e.g., candles) sources.
REFERENCES
1. Traynor, G. W.; Anthon, D. W.; Hollowell, C. D., Atmospheric Environment, 1982, 16, pp
2979-2987.
2. Yamanaka, S.; Hirose, H.; Takada, S., Atmospheric Environment, 1979, Vol. 13, pp 407-412.
3. Cheng, Y. S.; Bechtold, W. E.; Yu, C. C.; Hung, I. F., Aerosol Science and Technology, 1995,
23, pp 271-278.
4. Fine, P. M.; Cass, G. R.; Simoneit, B. R. T., Environ. Set. & Techn., 1999, 33, pp 2352-2362.
5. Alphen, M., The Science of the Total Environment, 1999, 243/244, pp 53-65.
6. Standard Test Method for Determining Air Change in a Single Zone by Means of a Tracer Gas
Dilution; ASTM E741-95; American Society for Testing and Materials: Philadelphia, PA.
7. Guo, Z.; Mosley, R. B,; McBrian, J,; Fortmann, R., In Proceedings of the 2000 Engineering
Solutions to Indoor Air Quality Problems Symposium, Air & Waste Management Association:
Pittsburgh, PA, 2000.
8. Guo, Z.; Tichenor, B. A.; Krebs, K. A.; Roache, N. F., Considerations on revision of emissions
testing protocols, in Characterizing Sources of Indoor Air Pollution and Related Sink Effects,
ASTM STP 1287, B.A. Tichenor, Ed., American Society for Testing and Materials, 1996, pp 225-
236.
-------
NRMRL-RTP-P-534
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing
1. REPORT NO,
EPA/6007 A-00/061
2.
3. RE
4. TITLE AND SUBTITLE
Test Methods to Characterize Participate Matter
Emissions and Deposition Hates in a Research House
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7.AUTHOR
------- |